How long must I wait for validation for SETI tasks?

Message boards : Number crunching : How long must I wait for validation for SETI tasks?
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · Next

AuthorMessage
Ianab
Volunteer tester

Send message
Joined: 11 Jun 08
Posts: 732
Credit: 20,635,586
RAC: 5
New Zealand
Message 2015143 - Posted: 12 Oct 2019, 2:56:56 UTC - in response to Message 2014923.  


Wouldn't it be better if they prioritized the project's tasks reported by how many they have received from individual's computers, the more received the sooner they get validation? I realize that many of the casual users would become a bit disappointed and may drop out of the SETI project, but hey... wouldn't it end up becoming more efficient for the SETI project as a whole?


The science we are doing isn't "time sensitive". Some of the data is years old. It doesn't matter if it takes a week. or a month to get a result. There are slow machines, there are part time machines, there are cellphones as Raspberry Pis, people go away on holiday, machines die etc. The system copes with that.

The priority is getting valid results back from everyone that want's to contribute. More complicated work allocation and deadlines wouldn't increase the amount of work done in the end, and say you did start allocating short deadlines to high turnover machines. Say 5 days? Someone takes a 2 week holiday and shuts down their computer. Bam, all those 100+ work units have timed out and need to get resent. With the shortest current deadline being about 3 weeks, that covers most of those scenarios. The high performance machines that you want to give shorter deadlines are the ones returning results in a day or 2 anyway

And no matter what the deadline situation, one work unit will ALLWAYS get returned before the other. So it doesn't even fix the "pending" count, on average 1/2 the units you return will drop into Pending for some period.


Not having a dig at you, and it's fine to ask questions like this. It's just I think it's a solution looking for a problem, because the problem isn't really a problem at all.
ID: 2015143 · Report as offensive
Profile George Project Donor
Volunteer tester
Avatar

Send message
Joined: 23 Oct 17
Posts: 222
Credit: 2,597,521
RAC: 13
United States
Message 2015167 - Posted: 12 Oct 2019, 13:49:30 UTC - in response to Message 2015143.  

Not having a dig at you, and it's fine to ask questions like this. It's just I think it's a solution looking for a problem, because the problem isn't really a problem at all.

As I've said before... many times... I have a whole lot to learn.

Thanks for your input, and to all who replied.
George

ID: 2015167 · Report as offensive
CryptokiD
Avatar

Send message
Joined: 2 Dec 00
Posts: 150
Credit: 3,216,632
RAC: 0
United States
Message 2015173 - Posted: 12 Oct 2019, 15:42:03 UTC

This brings up an idea for a contest. What is the oldest waiting for validation status work unit anyone can find?
Keith said hes got one from wayyy back on may 6th and its on its 4th wingman possibly going on a 5th wingman. Care to update us Keith? Can anyone beat Keiths workunit?
ID: 2015173 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 2015174 - Posted: 12 Oct 2019, 16:03:12 UTC - in response to Message 2015173.  

That sounds like fun. Ok, here is my submission. https://setiathome.berkeley.edu/workunit.php?wuid=3463464227
My third wingman gets a last crack at it until 13 Oct 2:20 UTC. If he does not return it, it goes on to a fourth wingman. So should know by early evening today if the task gets moved further out.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 2015174 · Report as offensive
Profile Joseph Stateson Project Donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 309
Credit: 70,759,933
RAC: 3
United States
Message 2015178 - Posted: 12 Oct 2019, 16:35:14 UTC - in response to Message 2015173.  
Last modified: 12 Oct 2019, 17:06:47 UTC

This brings up an idea for a contest. What is the oldest waiting for validation status work unit anyone can find?
Keith said hes got one from wayyy back on may 6th and its on its 4th wingman possibly going on a 5th wingman. Care to update us Keith? Can anyone beat Keiths workunit?


My oldest is July 19. Not even close to May.
https://setiathome.berkeley.edu/workunit.php?wuid=3569252203

Except for a few 100 that I recently downloaded, all my pending are from the WOW event. With cooler weather coming I will start heating house by firing up the SETI rig.
ID: 2015178 · Report as offensive
Profile Joseph Stateson Project Donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 309
Credit: 70,759,933
RAC: 3
United States
Message 2015180 - Posted: 12 Oct 2019, 16:40:37 UTC - in response to Message 2015139.  

I think before the next WOW event i need to put my experimental 16 x 3070 mining auto bunkering host running the multi spoofed new client working or i will be doomed


Auto bunkering is not that far fetched. I have come up with a scheme that can do it and will run some tests. Not going to give away any secrets but I might trade for secret inside info from the GPU Users Group.
ID: 2015180 · Report as offensive
Profile betreger Project Donor
Avatar

Send message
Joined: 29 Jun 99
Posts: 11360
Credit: 29,581,041
RAC: 66
United States
Message 2015183 - Posted: 12 Oct 2019, 17:02:34 UTC

My oldest = 25 Jul 2019, 10:28:47 UTC
ID: 2015183 · Report as offensive
juan BFP Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 16 Mar 07
Posts: 9786
Credit: 572,710,851
RAC: 3,799
Panama
Message 2015187 - Posted: 12 Oct 2019, 17:48:10 UTC - in response to Message 2015180.  
Last modified: 12 Oct 2019, 18:05:16 UTC

I think before the next WOW event i need to put my experimental 16 x 3070 mining auto bunkering host running the multi spoofed new client working or i will be doomed


Auto bunkering is not that far fetched. I have come up with a scheme that can do it and will run some tests. Not going to give away any secrets but I might trade for secret inside info from the GPU Users Group.

My host already runs on autobunkering+spoofing for the WoW event. But be aware, you have a big chance to crash some of them if anything goes wrong.
Now i just need to negotiate with my wife to allow me to build a Frankenstein like big mining host, she not even like my host on a TT X9 Cube, now imagine something even weird . That is a lot hard to do than code anything, sure you know what i mean. LOL

BTW There is a common mistake about our team, we not have any secret information or trade secrets, the code we use is exactly the same available to anyone on the Boinc site. We just care about the project impact of any new development, that is why we keep some closed to avoid any trouble with the servers load, but that is for the safe of the project not for any personal gain.

My old pending is https://setiathome.berkeley.edu/result.php?resultid=7810642481 Created on 28 Jun 2019, 18:04:42 UTC
ID: 2015187 · Report as offensive
Lazydude
Volunteer tester

Send message
Joined: 17 Jan 01
Posts: 45
Credit: 96,158,001
RAC: 136
Sweden
Message 2015189 - Posted: 12 Oct 2019, 18:01:46 UTC - in response to Message 2015183.  

https://setiathome.berkeley.edu/result.php?resultid=7853016842
12 July

Not cleared from dbase
https://setiathome.berkeley.edu/result.php?resultid=4297535947
26 Sep 2015, 5:59:35 UTC SETI@home v7
Anonymous platform (NVIDIA GPU)[url]
ID: 2015189 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 2015191 - Posted: 12 Oct 2019, 18:13:00 UTC

My third wingman gets a last crack at it until 13 Oct 2:20 UTC. If he does not return it, it goes on to a fourth wingman. So should know by early evening today if the task gets moved further out.

Almost guaranteed the task gets resent to a 4th wingman. The last wingman hasn't contacted the project since the day he was sent the task. All his tasks have timed out. So let's hope the _4 iteration is the lucky charm.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 2015191 · Report as offensive
juan BFP Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 16 Mar 07
Posts: 9786
Credit: 572,710,851
RAC: 3,799
Panama
Message 2015192 - Posted: 12 Oct 2019, 18:16:48 UTC - in response to Message 2015191.  

My third wingman gets a last crack at it until 13 Oct 2:20 UTC. If he does not return it, it goes on to a fourth wingman. So should know by early evening today if the task gets moved further out.

Almost guaranteed the task gets resent to a 4th wingman. The last wingman hasn't contacted the project since the day he was sent the task. All his tasks have timed out. So let's hope the _4 iteration is the lucky charm.

I always think on something, why not send the resent units to the fastest returning hosts, sure that could help to reduce the pendings list of all users and maybe could help to reduce a little the DB size.
Just an idea to add to the wishes list.
ID: 2015192 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 2015193 - Posted: 12 Oct 2019, 18:24:07 UTC - in response to Message 2015192.  

Yes, that would be helpful. But it would mean adding more server code. Where do you keep the database list of fastest returning hosts? What mechanism looks at the task resend count and then adds them to the pool of timed out tasks for the fastest hosts to pull from? What piece of code determines when a fast host is contacting the scheduler for more work and then directs it to the timed out pool first?
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 2015193 · Report as offensive
juan BFP Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 16 Mar 07
Posts: 9786
Credit: 572,710,851
RAC: 3,799
Panama
Message 2015194 - Posted: 12 Oct 2019, 18:36:39 UTC - in response to Message 2015193.  
Last modified: 12 Oct 2019, 18:39:44 UTC

Yes, that would be helpful. But it would mean adding more server code. Where do you keep the database list of fastest returning hosts? What mechanism looks at the task resend count and then adds them to the pool of timed out tasks for the fastest hosts to pull from? What piece of code determines when a fast host is contacting the scheduler for more work and then directs it to the timed out pool first?

They already have the Average turnaround time for each host. That could be used to set that. Sure need some coding from the server side but nothing extraordinary. You could complicate the code to infinite but i was talking about something simple like, send the resent only to host who has the Average turnaround time of less than 1 or 2 days, a simple "if", could do the task. Still the possibility of one off them not end the work but then it will be resent to another fast host, so the pendings will be cleared a lot faster than now. In a couple of days not in a couple of months like now.
ID: 2015194 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 2015195 - Posted: 12 Oct 2019, 18:45:40 UTC

But case in point. Look at my latest wingman. He has a 0.68 day turnaround. So he would qualify for your fast host criteria. But in fact he got 20 tasks once and never contacted the project ever again. Perfect definition of the one-time "drive-by" host that stumbles upon a mention of the project, signs up and then never runs the software or deletes the application and orphans the tasks on the host.
I would like these hosts to be eliminated first, then no issues of timed out tasks.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 2015195 · Report as offensive
juan BFP Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 16 Mar 07
Posts: 9786
Credit: 572,710,851
RAC: 3,799
Panama
Message 2015197 - Posted: 12 Oct 2019, 19:12:32 UTC - in response to Message 2015195.  

But case in point. Look at my latest wingman. He has a 0.68 day turnaround. So he would qualify for your fast host criteria. But in fact he got 20 tasks once and never contacted the project ever again. Perfect definition of the one-time "drive-by" host that stumbles upon a mention of the project, signs up and then never runs the software or deletes the application and orphans the tasks on the host.
I would like these hosts to be eliminated first, then no issues of timed out tasks.

I not said my idea was perfect, just simple and could help with most of the pendings, but that problem you pointed could be easy solved by adding another set of "if", resents only to hosts who crunched more than x number of WU and last contact was in the last x days.
As i said you could complicate the code changes up to infinite and solve every path, but we need keep the changes smaller as possible to be easy to coding.
ID: 2015197 · Report as offensive
Profile betreger Project Donor
Avatar

Send message
Joined: 29 Jun 99
Posts: 11360
Credit: 29,581,041
RAC: 66
United States
Message 2015200 - Posted: 12 Oct 2019, 19:34:26 UTC - in response to Message 2015197.  

But case in point. Look at my latest wingman. He has a 0.68 day turnaround. So he would qualify for your fast host criteria. But in fact he got 20 tasks once and never contacted the project ever again. Perfect definition of the one-time "drive-by" host that stumbles upon a mention of the project, signs up and then never runs the software or deletes the application and orphans the tasks on the host.
I would like these hosts to be eliminated first, then no issues of timed out tasks.

I not said my idea was perfect, just simple and could help with most of the pendings, but that problem you pointed could be easy solved by adding another set of "if", resents only to hosts who crunched more than x number of WU and last contact was in the last x days.
As i said you could complicate the code changes up to infinite and solve every path, but we need keep the changes smaller as possible to be easy to coding.

In that case since he isn't contacting the project he will not get the resend.
ID: 2015200 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 2015212 - Posted: 12 Oct 2019, 21:50:01 UTC

I've got 2 extremely old ones.
14th July- not due to be resent till the 29th Oct.
10th of July- not due to be resent till the 26th Oct.
Grant
Darwin NT
ID: 2015212 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 2015245 - Posted: 13 Oct 2019, 2:47:41 UTC

Well as expected my oldest task is going to go out to a fourth wingman with a deadline 7 weeks from today. Task is currently unsent but expect it to ship out today. The fast outage might have delayed it a bit.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 2015245 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 2015345 - Posted: 13 Oct 2019, 21:05:20 UTC

I had a quick look at mine when this conversation first got started, but had nothing earlier than late July - with the first significant batch coinciding with the start of the WOW! event. One I looked at seemed to be paired with a wingmate with a number of ghost tasks - the partner had a number of timeouts, but had completed and validated more recent tasks.

Today, I see I've received more credits than on any day since the final day of WOW! - http://boincstats.com/stats/0/user/detail/5509/charts. Anybody else seeing a lot of pendings validate in the last 24 hours?
ID: 2015345 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 5204
Credit: 840,779,836
RAC: 2,768
United States
Message 2015346 - Posted: 13 Oct 2019, 21:21:58 UTC - in response to Message 2015345.  

ID: 2015346 · Report as offensive
Previous · 1 · 2 · 3 · Next

Message boards : Number crunching : How long must I wait for validation for SETI tasks?


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.